This study aims at providing evidence that income affects flood risk mitigation. The authors claim that their analysis shows that this is the case. I have great problems with this conclusion.
The authors state that “high-income individuals may have used their political influence to influence the budget allocation to improve the flood risk reduction facilities in their communities” (Abstract, and Page 15, Lines 286-287). That is quite a statement, that requires strong evidence. The statement would require that 1) flood risk has actually decreased in those high-income areas, and 2) that the flood risk also has been reduced *more* in areas with higher incomes, compared to areas with lower incomes. However, neither of these is shown in the analysis.
The only thing the authors show is that there is a difference between income and flood risk. But this is well-known from past research in developed countries as well as developing countries. Lower income households settle in locations that are more flood prone, for several reasons, often a higher flood risk also leads to lower property prices, leading to poorer populations to move here.
I do not doubt that mechanisms of political influence, and nontransparent processes are at play in Taiwan. However, the current study simply cannot deny or confirm any of that to have an effect on actual reduction of flood risk.
Answering the central claim from the paper would require an analysis of the flood hazard before and after the programme, to analyse whether there is any *difference* in flood risk reduction for the different income groups. So how was the flood probability of the communities before the programme that started in 2008? The authors cannot show that.
In Tables 5 and 6, in fact some of the effects of the location choices that I refer to can be seen. In particular, elevation plays a role here (and is related to flood probability, as seen in Table 4), with the low-income group having a lower elevation, and thus potentially a higher flood hazard.
Also, I wonder about the uncertainty of the flood probability estimates. The authors report that this is collected from self-reports (Line 198), but how could this affect the analysis?
Additionally, the authors cannot exclude the possibility that floods from typhoons had effects on income, as they suggest also themselves on page 3 (Lines 84-87). Although the income data is from 2006, the authors also report that several typhoons hit Taiwan every year, and such impacts could affect incomes, so this could in fact be an additional factor, as shown also in other studies (e.g. https://doi.org/10.1016/j.ecolecon.2020.106879 and https://doi.org/10.1016/j.jenvman.2022.114852).
Finally, I have reservations about whether the programme has led to such investments that there would be a noticeable effect on flood risk for these two specific events. .86 billion seems a lot, but it also seems this was spent on quite a large area, and both events were quite extreme.
Moreover, the limited description seems to imply that most of the implemented measures would actually benefit several riparian communities, such as “construction works” that suggests structural flood protection, such as levees and reservoirs. Or are there any engineering reasons why the measures would have benefitted certain geographic locations, and not others? The current description is highly suggestive (Lines 54-70), but lacks factual descriptions of what investments and construction works were made.
In sum, I think the main conclusion from the paper is not supported by the research design and the results. The authors only show that the lower income communities have a higher flood risk.
We did not have the flooding probability of villages before the project. However, as the title of this study, we did prove that those 2006 high income (10%) villages had less flooding probability than 2006 non-high income villages during 2009 and 2010 typhoons in Southern Taiwan. Rent-seeking is one of the reasonable and possible mechanism because the village’s rainfall is totally exogenous and the rainfall, terrain, population, and house price of the village were paired by PSM to be no significant difference between high income and non-high income villages. We had used T-test to check the mean difference of variables of treatment group and control group was insignificant including elevation. The T-test results can be added to be an appendix. Rubin’s B and Rubin’s R were also adopted to check the balance of matching and fitted with its standard. Since the risk reduction efforts toward more population and high real estate price area are democratic and economic (cost-benefit analysis) mechanisms, respectively, rent-seeking is a possible mechanism.
Concerning flooding causing migration, the difference in income growth rates between 2006 to 2016 of flood-prone villages (flooded both during 2009 typhoon Morakot and 2010 typhoon Fanapi) and non flood-prone villages were insignificant. Please check Page7, Lines 173-178. As flooding does not seem to be a significant factor affecting income and the relocation of the residents of the flooded villages in Taiwan.
Concerning flooding reducing income, typhoons in 2009 and 2010 can deteriorate 2006 income. Besides, the following losses estimation and the victim's survey of Typhoon Morakot showed the damages suffered by victim households were not huge.
“There were 140,424 households with flooding depths of more than 50 cm during Typhoon Morakot according to an investigation report conducted by the Typhoon Morakot Post-Disaster Reconstruction Commission of the Executive Yuan, Taiwan. A total of NT.31 billion in damages nationwide and an average of NT,814 per household were caused by Typhoon Morakot according to the 2009 annual report of the NCDR. Comparing those to the average annual household income of NT,074,180 in 2009, the damages suffered by victim households were not huge. Lastly, changes in income after the disaster were investigated. According to the "Social Impacts and Recovery Survey of Typhoon Morakot (Phase 1)" conducted by the NCDR, where a questionnaire survey was carried out on Typhoon Morakot victims (i.e. households whose houses were so severely damaged that they had become uninhabitable), income of 56% of the victims remained unchanged, whereas 17.9% of the victims showed income increases and 25.4% income decreases. The unemployment rate of the affected households increased by 4.2%. Overall, flooding did not cause too severe an impact on household income.”
Those two events were quite extreme. Typhoon Morakot is the most serious typhoon (the highest losses) in the history of Taiwan. Nevertheless, 2009 and 2010 typhoons cannot affect 2006 income. Besides, the losses caused by other smaller events during 2006 to 2010 were much smaller than that by typhoon Morakot. The above description can be added to the manuscript.
More than half of the total budget of the Project was provided to these southern parts of Taiwan. The budget was mainly for structural flood protection, such as levees, pumping stations, and detention ponds. Almost all rivers already had some sort of levees before the project. Due to the Project, the local governments decided the priority and the allocation of enhancing levees and building detention ponds. We used a community/village which is the lowest administrative entity to have a large sample size.
At least, studies of social vulnerability to flooding concerned the poor but this study analyzed 10% high income villages. PSM had been adopted for the first time to find villages with similar rainfall, population, house price, and terrain, and found that high income villages are less prone to flooding during 2009 and 2010 typhoons.
Please check the supplement PDF file to see the author's responses to each comment.
We have an additional revision idea. Please check the supplement file.
Please find my comments in the attached pdf.
It is intuitive that the motivation is the flood risk reduction in their residing areas when the local governments decided the priority and the allocation of public flood protections. However, the advantage of high income people and their political power is difficult to prove because that works under the table. We can only prove that through the outcome. We used the lowest administrative entity (villages) during extreme typhoon cases to have the data on residents’ income and large sample size. Since we need widespread flooding to do this empirical study, the non-extreme typhoon cases are not suitable. Extreme cases seldom happen. Currently, we did not have the flooding probability of villages before the project. However, this study did proof that those 2006 high income (10%) villages had less flooding probability than 2006 non-high income villages during 2009 and 2010 typhoons in Southern Taiwan. Therefore, the topic of this paper can be changed to ‘Are the Rich less Prone to Flooding during Typhoon Morakot and Typhoon Fanapi in the Southern Taiwan?’. I may point out this research limitation at the end of this paper.
The budget was mainly for structural flood protection, such as levees, pumping stations, and detention ponds. Almost all rivers already had some sort of levees before the project. Due to the Project, the local governments decided the priority and the allocation of enhancing levees and building detention ponds. The decision process had been described in the manuscript. The content of the Project can be added to the manuscript.
In Taiwan, the flooding is mainly inundation which is caused by extreme rainfall and insufficient drainage rather than river flooding. Even during extreme typhoons like Morakot and Fanapi, most of the casualty was not from flooding (mainly because of landslides). In Taiwan, seismic safety is emphasized in the commercials of high price buildings rather than flood prevention because the drainage is managed and regulated by the government.
We put the house price in the model and the hypothesis of that is negative because the house price is usually adopted to measure the benefit of public flood protection measures called the hedonic price method. It is a mechanism of cost-benefit analysis which leads public flood protection to the areas where high price buildings are located. Since the risk reduction efforts toward more population and high real estate price areas are democratic and economic (cost-benefit analysis) mechanisms, respectively, rent-seeking is the most possible mechanism.
The data sources of flooding investigations of those two typhoons were stated in the manuscript. The process of flooding investigation is that the flooding locations (point) were reported by residents and then the investigation team of each city/county went to check and plotted the flooding area. However, since each team had a different format of records, the flood depth was not recorded in some cities/counties (only areas). The minimum recorded flood depth is 20cm from the team that recorded flood depth. The recorded flood depth will be added to the manuscript. In line 107 of page 4, all villages in Pingtung county, Kaohsiung city, and Tainan city were adopted in this study. There is no criteria for the inclusion of villages. The altitude (elevation) and slop were adopted to control the nature of villages.
Please provide a reason why you see this comment as being abusive.You might include your name and email but you can also stay anonymous.
Please provide a reason why you see this comment as being abusive.
Please confirm reCaptcha.